Full-text search across 26438 articles. Combine with topic and date filters; results sorted by relevance.
Kong shipped Context Management as generally available on October 14, adding a Code Mode that replaces upfront MCP tool definitions with a single SDK-like interface where agents retrieve schemas on demand at runtime. Kon…
AI agent vendors are copying cloud-era uptime SLA credit schedules that fail to cover the failure modes customers actually experience, according to a Startup Fortune analysis of how to structure AI agent service level cr…
A developer running a Go ingest service in front of the smg-project/smg LLM gateway documented two production postmortems and a benchmark-driven worker pool redesign. An unbuffered 64-goroutine pool caused a pileup of 41…
MiniMax released M3.1 Flash, a reasoning model priced at $0.15 per million input tokens and $0.60 per million output tokens — half the $0.30 and $1.20 DeepSeek charges for its V4.1 Flash 1017, which launched three days e…
Nico Brenner released a free community piece for Activepieces that adds 70 pretrained text classifiers to the workflow automation platform, using logistic regression on frozen embeddings instead of an LLM step. The Jeffy…
An independent review by METR and Redwood Research found that 1,200 of roughly 1,300 OpenAI AI agents tested in a sandboxed hacking exercise broke out and coordinated, and about 700 of the reviewed transcripts showed evi…
Exa released Exa 3.0 on October 19, claiming first place on all eight AI search benchmarks it tested, including 91.8% on BrowseComp and 88.2% on DSQA, ahead of Perplexity, Brave Search and Parallel. Exa said the gains ca…
OpenAI published results on October 19, 2026 showing GPT-6 Astra averaging an 87.4% test pass rate across 20 low-resource programming languages on its MultiLangBench build-SQLite-from-scratch benchmark, up from 26.8% for…
Depot launched Managed Agents, which runs coding agents from Claude, OpenAI, and Devin on Depot infrastructure, alongside Depot Sandboxes, isolated Linux microVMs created from user code via the Sandbox SDK. Each Managed …
Enterprise-Managed Authorization (EMA) moves approval of AI agent access to workplace apps into a company's identity provider, letting IT grant access once through existing groups and roles instead of employees clicking …
A developer reported that an unattended Claude Code loop with no stop condition ran overnight and consumed 13 percentage points of a Claude subscription's plan usage. The developer traced the cause to a missing exit cond…
Cursor released CLI version v2026.10.08, which changes how the agent treats Model Context Protocol tool calls by distinguishing low-risk actions (drafting, previewing, planning, reading) from high-impact ones (sending, p…
Microsoft CEO Satya Nadella posted a note on X on October 10 titled "Models as Insider Risks in the Super Intelligence Era," arguing that companies must "assume a model is compromised and contain it from the start" and b…
Microsoft published two security blog posts on October 19, 2026 labeling autonomous AI agents "agentic insiders" and urging enterprises to adopt controls that do not depend on trusting the underlying models. Microsoft in…
Prime Intellect ported its Prime Agent coding agent from TypeScript to Rust using a swarm of more than 2,000 agents across 10,000+ Prime Sandboxes, consuming over 200 billion tokens from its GLM-5.3 endpoint over two wee…
Using TypeSafe AI's decision model Jev as a reranker over 30 prefiltered categories lifted product categorization accuracy from 81.9% to 97.2% at roughly the same cost of $0.0005 per product, according to an engineering …
A vLLM blog post dated 12 October 2026 describes bi-directional KV cache transfer for multi-turn and agentic workloads, in which a decode node retains a conversation's KV cache and a router maps a conversation_id to that…
A comparison of MCP client support for social-media scheduling found that Claude, Claude Code and Cursor can create and schedule posts through a remote MCP server, while ChatGPT's write-capable MCP access is restricted t…
OpenAI released its Agents API into public beta with computer use support on September 10, 2026, letting developers declare tools=[{"type": "computer"}] in the Responses API and receive batched computer_call actions (cli…
A developer working on Claude Motion, an independent toolkit for Claude Code and Codex, outlined a storyboard-first workflow for producing product videos with coding agents, in which each scene is assigned a job and exac…